Time Series Analysis

Catalog description

Time series analysis is concerned with experimental data that have been observed at different points in time. Samples taken from such data over a sequence of time periods often show that adjacent observations are not always independent. Hence, the usual techniques from classical statistics, developed primarily for independent, identically distributed observations, are not applicable. The objectives of this course are to provide tools and methods to describe important features of time-series, such as trend, seasonality, correlated errors and periodicity; to apply the commonly used statistical and computational techniques to analyze data and make inferences, such as estimation and forecasts; to understand time series models and their properties; to fit such models to experimental data. The methodology will be illustrated with the analysis of different data sets arising from economics, biology, medicine, and engineering, etc.